Abstract: Cardiovascular disease is the leading cause of death among postmenopausal women. Growing evidence exists that HRT may reduce cardiovascular disease and mortality. This has led to the use of postmenopausal HRT for primary prevention of heart disease, however it remains unresolved which women are likely to benefit most from this clinical approach. This is a complicated question because many potential mechanisms are involved by which hormone therapy (HRT) may confer cardioprotection. The Postmenopausal Estrogen/Progestins Intervention (PEPI) trial examined the relative impact of four regimens of hormone therapy on a range of cardiovascular disease risk factors (e.g. lipids, blood pressure, insulin/glucose, hemostasis factors). The trial was well conducted; its data have been thoroughly edited and have yielded a series of important publications. To date, however, publications have focused on average effects without a thorough exploration of the range of, and interplay among, the various effects across the 875 enrolled in the trial. We propose to extend the analysis of PEPI data: 1) characterize the distributions of responses to hormone therapy with respect to risk factors for cardiovascular disease (lipids/lipoproteins, blood pressure, insulin/glucose, hemostasis factors); 2) examine the multivariate patterns of treatment effects among these cardiovascular risk factors and on other outcomes (symptomatology, bone, endometrial); 3) examine clinical and demographic factors that may affect these relationships and, in doing so, characterize women who may vary with respect to their "sensitivity" to the separate effects of estrogen and progestin therapy; 4) examine closely patterns of adherence and their relationship to response. An experienced team of statisticians/epidemiologists is proposed to conduct and publish results from these analyses. Its members are thoroughly familiar with the data from the PEPI trial. The results of the proposed analyses are expected to hold considerable clinical significance and may help to interpret emerging findings from event-based trials.